Final project information

Report guidelines
Demonstrations & due dates
Robot signup
Maximal information
Amos Eaton first floor hallway map
Finding research papers

Final project report guidelines

I suggest having the basic organization of a conference paper: introduction, related work, the body, and conclusions.
There are no length requirements. Don't make it so short that you omit important details and results, but don't making it too long by including unimportant details. Use a single column and single spaced format with reasonable (> 1 inch) margins.
Pretend I've forgotten all about what you're doing for your project, so start from the beginning, e.g. "We're combining Markov localization and motion planning to make the robot..."
Your report should describe what you did, how you did it, and what the results were. Be detailed enough to be technically precise (e.g. exactly how did you do what you did), but not so detailed as to include the mundane (e.g., don't give me a detailed line by line description of your code).
Don't give me a blog, i.e., don't give me a step by step account of your project: "first we wrote some code to do this and then we compiled it and tested it on the robot and then we wrote some more code..."
Feel free to describe things that you attempted but did not end up using in the final version of your code.
You can use your reading reports in the related work section, but you should edit them down --- a single conference paper should only end up being about two paragraphs or so.

Demonstrations & due dates

I am available for final project demonstrations
- Friday April 30: 10-12, 1-5
- Sunday May 2: 2-5
- Monday May 3: 9-10, 11-6
- Tuesday May 4: 9-12
You may go ahead and sign up for 1 hour "demo" slots right away. The actual demo will not actually take that long (maybe 15 min?) but this will give you time to get set up.
The demo does not count towards your "two hours of allowed reserved time" per day. However, you cannot reserve other times on these days until 48 hours earlier, as per the regular signup policy.
I must have final project reports by noon on Tuesday May 4. You can drop them off at my office (slip it under the door if I am not there). This is a hard deadline.

Robot signup

This is just what we have discussed in class:

Each final project team may reserve upto 2 hours of time on the robot per day (using the yahoo calendar account), no more than 48 hours in advance.
Any team may use unreserved time on the robot.
If your team has been using more than 2 hours of unreserved time (on a given day) and another team comes in for unreserved time, you must yield to them.
If the robot battery voltage drops below 20 volts, keep it plugged into the charger. (When you finish using the robot, follow the regular charging policy.)
Please be reasonable about reserving time or cancelling a reservation in advance. (I don't think this has been a problem in previous years, but we can make up some rules to cover this case if it becomes necessary.)

Maximal information

I have installed CGAL, dolt, SVL, and filereader on maximal.robotics
Everything is in /local/scratch0/ira
Be careful that you do not use the old versions of these libraries in /usr/local They will probably not quite compile with this year's support code from the assignments.
To use these libraries, you will need to set:
```
export DOLT_MAKEFILE=/local/scratch0/ira/make/Makefile-dolt
export CGAL_MAKEFILE=/local/scratch0/ira/CGAL-3.0/make/makefile_i686_Linux-2.2.14-5.0_g++-2.95
```
Since you have a local directory on this machine, you should go ahead and put the above lines in your .bashrc file. (Since it's a local directory, this will not mess up your regular CS environment.)
Note that maximal has only g++ version 2.95, whereas we have been using version 3.3. However, I don't think this should cause problems — I was able to compile my assignment 2 and 3 solutions without any problem.

Amos Eaton first floor hallway map

Here's the data: ae-map.txt
This file is just a series of points (representing corners). The points constitute a polygon. You can just feed them into a dolt drawablePolygon.
All measurements are in meters.
The first point in the map (which is the origin) is in the far hallway near the restrooms, i.e. just past the door of room 123A. It's essentially the mirror image of the corner to the left of the lab door (from standing in the hallway).
The map includes the main AE corridor (which wraps around the entire first floor) and includes the lounge, with one door closed. You may want to remove the lounge data and close off its doorway.
The measurements should be reasonably accurate (within a centimeter or two for each wall). The raw measurements were "embedded" in the plane using a spring model with a strict rectilinear assumption.
The corner in the hallway that is farthest away from the lab (the back right corner) has a door that is usually open at an angle. In the data, the door was blocked off to present a rectilinear corner.

Finding research papers

Reminder: Unless you get approval in advance, the research papers you review for the final project must be from one of the following journals:
- IEEE Transactions on Robotics and Automation
- International Journal of Robotics Research
- Autonomous Robots
- Robotics and Autonomous Systems
- Computational Geometry: Theory and Applications
(note that I've added the Autonomous Robots journal to the list, and I got the name the last one wrong on the final project info handout)
Papers may also be from one of the following conferences:
- IEEE International Conference on Robotics and Automation (ICRA)
- IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
You can papers from all but one of these sources online. Our library has electronic subscriptions to all the above publications except for "Computational Geometry: Theory and Applications" --- for this one, you'll have to go to the library because it's only available there in hardcopy.
For the IEEE publications, you can go to IEEExplore.
- For the Transactions on Robotics and Automation, go to "Journals and magazines" and search for "robotics and automation"
- For ICRA, go to "Conference Proceedings" and search for "Robotics and Automation"
- For IROS, go to "Conference Proceedings" and search for "Intelligent Robots"
You'll have to have an RPI IP address to get access to IEEExplore under the RPI subscription. (Or do the proxy thing...)
For the other journals available online, go to the library web page and do a search for the title.
Another source of papers is the authors' web sites. Practically all US authors will have their papers available there. [For the markov localization groups, you can find the AAAI paper on Sebastian Thrun's web page at Stanford.]
There are several ways to find other research papers on a topic you are interested in:
- Use the INSPEC database, available through the library web page.
  Go to "Resources and Collections", then "Databases", then "Engineering", and you'll find a link to the "INSPEC" database there.
- Try CiteSeer
- Search the web. Find web pages for researchers doing the work, and they will have lists of their publications (if not publications online).